AVP Gen2.0
(Reusable system)
New Progress！
During the internship in the deep blue aerospace. I developed a full scale software in loop control algorithm simulation.
The physical simulation is base on the actual rocket(XY-2 in development) full scale physical simulation. And the aerodynamics is considered.
In the rocket recovery control system, we implemented a convex optimization-based linear quadratic programming controller to determine rocket engine ignition timing for autonomous landing. The system controls engine startup timing and thrust output (constrained between 60%-100%), while accounting for engine response delays. The control objective is to land the rocket precisely within a 50m×50m designated area with touchdown velocity below 2m/s.
The convex optimization control demonstrates excellent performance. Simulations show that initial position errors up to 700 meters can be corrected, achieving landing accuracy within 15 meters – significantly outperforming alternative control methods like ADRC.
The following video demonstrates the complete software-in-the-loop simulation process.
And this one is using ADRC controller.
Why AVP Gen2.0
With the basis of reusable rocket AVP Gen1.0, in order to build a more powerful VTVL system to support my following star ship and missile prototype, AVP Gen2.0 come.
In AVP Gen2.0, I am more focusing on the simulation, which I build a full physical simulation platform and flight control algorism system and it’s code generation system so the I can achieve the rapid prototyping as fast, as accurate as possible.
So far the full size star ship physical simulation and it’s control algorism testing is finished.
Wait for further progress….